Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1253956 > unrolled thread

[RFC][PATCH 0/4] tracing: Add set_event_pid to filter on pids for events

Back to article view | Back to linux.kernel

Contents

  [RFC][PATCH 0/4] tracing: Add set_event_pid to filter on pids for events Steven Rostedt <rostedt@goodmis.org> - 2015-10-22 18:20 +0200
    [RFC][PATCH 3/4] tracing: Implement event pid filtering Steven Rostedt <rostedt@goodmis.org> - 2015-10-22 18:20 +0200
    [RFC][PATCH 2/4] tracing: Add set_event_pid directory for future use Steven Rostedt <rostedt@goodmis.org> - 2015-10-22 18:20 +0200

#1253956 — [RFC][PATCH 0/4] tracing: Add set_event_pid to filter on pids for events

Currently the only way to filter pids for events is to add it to the
filter file in the tracefs directory. Something like:

  echo 'common_pid == 2123 || common_pid == 4211' > filter

The problem with this is that there's a very limited size that the
filter can accept. Thus, trying to trace all processes of hackbench
is not going to cut it.

I've added a new file called set_event_pid which when not empty will
filter the events if the event is for one of the listed pids.
It adds special hooks for sched_switch and sched_wakeup to handle
next_pid and pid fields respectively.

I have code that will extend this to also dynamically add the children
on these pids to the filter on fork. But that's still being worked on.

The filtering is done by a flag being set at sched_switch time,
by hooking into the sched_switch tracepoint. When a process is scheduled in,
it is checked against the list of pids in the set_event_pid file (bsearch)
and if the files is not empty and the pid does not exist, a flag is
set to ignore tracing the current task. Then all other tracepoints only
need to check the flag to see if it should be ignored. This only affects
events within a specific tracing instance. perf and other tracing instances
will not be affected by another instance filtering.

The first thing that needed to be done was to add a priority to callbacks
of tracepoints. The hook added to sched_switch that determines the
if the task should be filtered or not must execute before any other
callbacks for that tracepoint. Otherwise filtering the sched_switch
tracepoint itself would not work.

The list of pids is allocated by pages. When changing (adding, removing
or clearing pids) the file special care must be used, as there are
no locks held when reading the list at sched switch. Normal RCU handling
is used (synchronize_sched) as all callbacks from tracepoints are
called with rcu_read_lock_sched() held.

This will allow for much better task filtering in the future.

Enjoy,

-- Steve



This series can be found at:

  git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace.git
ftrace/event-pid-filter

Head SHA1: d4f2f260e337242f8a5773f48b98463b95def9a9


Steven Rostedt (Red Hat) (4):
      tracepoint: Give priority to probes of tracepoints
      tracing: Add set_event_pid directory for future use
      tracing: Implement event pid filtering
      tracing: Check all tasks on each CPU when filtering pids

----
 include/linux/trace_events.h |   7 +
 include/linux/tracepoint.h   |  13 ++
 kernel/trace/trace.h         |   9 +
 kernel/trace/trace_events.c  | 451 +++++++++++++++++++++++++++++++++++++++++++
 kernel/tracepoint.c          |  61 ++++--
 5 files changed, 530 insertions(+), 11 deletions(-)
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [next] | [standalone]

#1253959 — [RFC][PATCH 3/4] tracing: Implement event pid filtering

From: "Steven Rostedt (Red Hat)" <rostedt@goodmis.org>

Add the necessary hooks to use the pids loaded in set_event_pid to filter
all the events enabled in the tracing instance that match the pids listed.

Two probes are added to both sched_switch and sched_wakeup tracepoints to be
called before other probes are called and after the other probes are called.
The first is used to set the necessary flags to let the probes know to test
if they should be traced or not.

The sched_switch pre probe will set the "ignore_pid" flag if neither the
previous or next task has a matching pid.

The sched_switch probe will set the "ignore_pid" flag if the next task
does not match the matching pid.

The pre probe allows for probes tracing sched_switch to be traced if
necessary.

The sched_wakeup pre probe will set the "ignore_pid" flag if neither the
current task nor the wakee task has a matching pid.

The sched_wakeup post probe will set the "ignore_pid" flag if the current
task does not have a matching pid.

Cc: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
---
 include/linux/trace_events.h |   7 ++
 kernel/trace/trace.h         |   2 +
 kernel/trace/trace_events.c  | 148 ++++++++++++++++++++++++++++++++++++++++++-
 3 files changed, 154 insertions(+), 3 deletions(-)

diff --git a/include/linux/trace_events.h b/include/linux/trace_events.h
index f85693bbcdc3..429fdfc3baf5 100644
--- a/include/linux/trace_events.h
+++ b/include/linux/trace_events.h
@@ -328,6 +328,7 @@ enum {
 	EVENT_FILE_FL_SOFT_DISABLED_BIT,
 	EVENT_FILE_FL_TRIGGER_MODE_BIT,
 	EVENT_FILE_FL_TRIGGER_COND_BIT,
+	EVENT_FILE_FL_PID_FILTER_BIT,
 };
 
 /*
@@ -341,6 +342,7 @@ enum {
  *                   tracepoint may be enabled)
  *  TRIGGER_MODE  - When set, invoke the triggers associated with the event
  *  TRIGGER_COND  - When set, one or more triggers has an associated filter
+ *  PID_FILTER    - When set, the event is filtered based on pid
  */
 enum {
 	EVENT_FILE_FL_ENABLED		= (1 << EVENT_FILE_FL_ENABLED_BIT),
@@ -351,6 +353,7 @@ enum {
 	EVENT_FILE_FL_SOFT_DISABLED	= (1 << EVENT_FILE_FL_SOFT_DISABLED_BIT),
 	EVENT_FILE_FL_TRIGGER_MODE	= (1 << EVENT_FILE_FL_TRIGGER_MODE_BIT),
 	EVENT_FILE_FL_TRIGGER_COND	= (1 << EVENT_FILE_FL_TRIGGER_COND_BIT),
+	EVENT_FILE_FL_PID_FILTER	= (1 << EVENT_FILE_FL_PID_FILTER_BIT),
 };
 
 struct trace_event_file {
@@ -429,6 +432,8 @@ extern enum event_trigger_type event_triggers_call(struct trace_event_file *file
 extern void event_triggers_post_call(struct trace_event_file *file,
 				     enum event_trigger_type tt);
 
+bool trace_event_ignore_this_pid(struct trace_event_file *trace_file);
+
 /**
  * trace_trigger_soft_disabled - do triggers and test if soft disabled
  * @file: The file pointer of the event to test
@@ -448,6 +453,8 @@ trace_trigger_soft_disabled(struct trace_event_file *file)
 			event_triggers_call(file, NULL);
 		if (eflags & EVENT_FILE_FL_SOFT_DISABLED)
 			return true;
+		if (eflags & EVENT_FILE_FL_PID_FILTER)
+			return trace_event_ignore_this_pid(file);
 	}
 	return false;
 }
diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h
index 250481043bb5..89ffdaf3e371 100644
--- a/kernel/trace/trace.h
+++ b/kernel/trace/trace.h
@@ -156,6 +156,8 @@ struct trace_array_cpu {
 	pid_t			pid;
 	kuid_t			uid;
 	char			comm[TASK_COMM_LEN];
+
+	bool			ignore_pid;
 };
 
 struct tracer;
diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c
index 2ad7014707ee..ab07058e27c1 100644
--- a/kernel/trace/trace_events.c
+++ b/kernel/trace/trace_events.c
@@ -22,6 +22,8 @@
 #include <linux/slab.h>
 #include <linux/delay.h>
 
+#include <trace/events/sched.h>
+
 #include <asm/setup.h>
 
 #include "trace_output.h"
@@ -212,12 +214,32 @@ int trace_event_raw_init(struct trace_event_call *call)
 }
 EXPORT_SYMBOL_GPL(trace_event_raw_init);
 
+bool trace_event_ignore_this_pid(struct trace_event_file *trace_file)
+{
+	struct trace_array *tr = trace_file->tr;
+	struct trace_array_cpu *data;
+	struct trace_pid_list *pid_list;
+
+	pid_list = rcu_dereference_sched(tr->filtered_pids);
+	if (!pid_list)
+		return false;
+
+	data = this_cpu_ptr(tr->trace_buffer.data);
+
+	return data->ignore_pid;
+}
+EXPORT_SYMBOL_GPL(trace_event_ignore_this_pid);
+
 void *trace_event_buffer_reserve(struct trace_event_buffer *fbuffer,
 				 struct trace_event_file *trace_file,
 				 unsigned long len)
 {
 	struct trace_event_call *event_call = trace_file->event_call;
 
+	if ((trace_file->flags & EVENT_FILE_FL_PID_FILTER) &&
+	    trace_event_ignore_this_pid(trace_file))
+		return NULL;
+
 	local_save_flags(fbuffer->flags);
 	fbuffer->pc = preempt_count();
 	fbuffer->trace_file = trace_file;
@@ -459,15 +481,114 @@ static int cmp_pid(const void *key, const void *elt)
 	return 1;
 }
 
+static bool
+check_ignore_pid(struct trace_pid_list *filtered_pids, struct task_struct *task)
+{
+	pid_t search_pid;
+	pid_t *pid;
+
+	/*
+	 * Return false, because if filtered_pids does not exist,
+	 * all pids are good to trace.
+	 */
+	if (!filtered_pids)
+		return false;
+
+	search_pid = task->pid;
+
+	pid = bsearch(&search_pid, filtered_pids->pids,
+		      filtered_pids->nr_pids, sizeof(pid_t),
+		      cmp_pid);
+	if (!pid)
+		return true;
+
+	return false;
+}
+
+static void
+event_filter_pid_sched_switch_probe_pre(void *data,
+		    struct task_struct *prev, struct task_struct *next)
+{
+	struct trace_array *tr = data;
+	struct trace_pid_list *pid_list;
+
+	pid_list = rcu_dereference_sched(tr->filtered_pids);
+
+	this_cpu_write(tr->trace_buffer.data->ignore_pid,
+		       check_ignore_pid(pid_list, prev) &&
+		       check_ignore_pid(pid_list, next));
+}
+
+static void
+event_filter_pid_sched_switch_probe_post(void *data,
+		    struct task_struct *prev, struct task_struct *next)
+{
+	struct trace_array *tr = data;
+	struct trace_pid_list *pid_list;
+
+	pid_list = rcu_dereference_sched(tr->filtered_pids);
+
+	this_cpu_write(tr->trace_buffer.data->ignore_pid,
+		       check_ignore_pid(pid_list, next));
+}
+
+static void
+event_filter_pid_sched_wakeup_probe_pre(void *data, struct task_struct *task)
+{
+	struct trace_array *tr = data;
+	struct trace_pid_list *pid_list;
+
+	/* Nothing to do if we are already tracing */
+	if (!this_cpu_read(tr->trace_buffer.data->ignore_pid))
+		return;
+
+	pid_list = rcu_dereference_sched(tr->filtered_pids);
+
+	this_cpu_write(tr->trace_buffer.data->ignore_pid,
+		       check_ignore_pid(pid_list, task));
+}
+
+static void
+event_filter_pid_sched_wakeup_probe_post(void *data, struct task_struct *task)
+{
+	struct trace_array *tr = data;
+	struct trace_pid_list *pid_list;
+
+	/* Nothing to do if we are not tracing */
+	if (this_cpu_read(tr->trace_buffer.data->ignore_pid))
+		return;
+
+	pid_list = rcu_dereference_sched(tr->filtered_pids);
+
+	/* Set tracing if current is enabled */
+	this_cpu_write(tr->trace_buffer.data->ignore_pid,
+		       check_ignore_pid(pid_list, current));
+}
+
 static void __ftrace_clear_event_pids(struct trace_array *tr)
 {
 	struct trace_pid_list *pid_list;
+	struct trace_event_file *file;
+	int cpu;
 
 	pid_list = rcu_dereference_protected(tr->filtered_pids,
 					     lockdep_is_held(&event_mutex));
 	if (!pid_list)
 		return;
 
+	unregister_trace_sched_switch(event_filter_pid_sched_switch_probe_pre, tr);
+	unregister_trace_sched_switch(event_filter_pid_sched_switch_probe_post, tr);
+
+	unregister_trace_sched_wakeup(event_filter_pid_sched_wakeup_probe_pre, tr);
+	unregister_trace_sched_wakeup(event_filter_pid_sched_wakeup_probe_post, tr);
+
+	list_for_each_entry(file, &tr->events, list) {
+		clear_bit(EVENT_FILE_FL_PID_FILTER_BIT, &file->flags);
+	}
+
+	for_each_possible_cpu(cpu)
+		per_cpu_ptr(tr->trace_buffer.data, cpu)->ignore_pid = false;
+
 	rcu_assign_pointer(tr->filtered_pids, NULL);
 
 	/* Wait till all users are no longer using pid filtering */
@@ -1429,13 +1550,14 @@ static int max_pids(struct trace_pid_list *pid_list)
 }
 
 static ssize_t
-ftrace_event_pid_write(struct file *file, const char __user *ubuf,
+ftrace_event_pid_write(struct file *filp, const char __user *ubuf,
 		       size_t cnt, loff_t *ppos)
 {
-	struct seq_file *m = file->private_data;
+	struct seq_file *m = filp->private_data;
 	struct trace_array *tr = m->private;
 	struct trace_pid_list *filtered_pids = NULL;
 	struct trace_pid_list *pid_list = NULL;
+	struct trace_event_file *file;
 	struct trace_parser parser;
 	unsigned long val;
 	loff_t this_pos;
@@ -1564,15 +1686,35 @@ ftrace_event_pid_write(struct file *file, const char __user *ubuf,
 
 	rcu_assign_pointer(tr->filtered_pids, pid_list);
 
-	mutex_unlock(&event_mutex);
+	list_for_each_entry(file, &tr->events, list) {
+		set_bit(EVENT_FILE_FL_PID_FILTER_BIT, &file->flags);
+	}
 
 	if (filtered_pids) {
 		synchronize_sched();
 
 		free_pages((unsigned long)filtered_pids->pids, filtered_pids->order);
 		kfree(filtered_pids);
+	} else {
+		/*
+		 * Register a probe that is called before all other probes
+		 * to set ignore_pid if next or prev do not match.
+		 * Register a probe this is called after all other probes
+		 * to only keep ignore_pid set if next pid matches.
+		 */
+		register_trace_prio_sched_switch(event_filter_pid_sched_switch_probe_pre,
+						 tr, INT_MAX);
+		register_trace_prio_sched_switch(event_filter_pid_sched_switch_probe_post,
+						 tr, 0);
+
+		register_trace_prio_sched_wakeup(event_filter_pid_sched_wakeup_probe_pre,
+						 tr, INT_MAX);
+		register_trace_prio_sched_wakeup(event_filter_pid_sched_wakeup_probe_post,
+						 tr, 0);
 	}
 
+	mutex_unlock(&event_mutex);
+
 	ret = read;
 	*ppos += read;
 
-- 
2.6.1


--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [prev] | [next] | [standalone]

#1253960 — [RFC][PATCH 2/4] tracing: Add set_event_pid directory for future use

From: "Steven Rostedt (Red Hat)" <rostedt@goodmis.org>

Create a tracing directory called set_event_pid, which currently has no
function, but will be used to filter all events for the tracing instance or
the pids that are added to the file.

The reason no functionality is added with this commit is that this commit
focuses on the creation and removal of the pids in a safe manner. And tests
can be made against this change to make sure things are correct before
hooking features to the list of pids.

Cc: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
---
 kernel/trace/trace.h        |   7 ++
 kernel/trace/trace_events.c | 287 ++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 294 insertions(+)

diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h
index fb8a61c710ea..250481043bb5 100644
--- a/kernel/trace/trace.h
+++ b/kernel/trace/trace.h
@@ -176,6 +176,12 @@ struct trace_options {
 	struct trace_option_dentry	*topts;
 };
 
+struct trace_pid_list {
+	unsigned int			nr_pids;
+	int				order;
+	pid_t				*pids;
+};
+
 /*
  * The trace array - an array of per-CPU trace arrays. This is the
  * highest level data structure that individual tracers deal with.
@@ -201,6 +207,7 @@ struct trace_array {
 	bool			allocated_snapshot;
 	unsigned long		max_latency;
 #endif
+	struct trace_pid_list	__rcu *filtered_pids;
 	/*
 	 * max_lock is used to protect the swapping of buffers
 	 * when taking a max snapshot. The buffers themselves are
diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c
index d120cfe3cca7..2ad7014707ee 100644
--- a/kernel/trace/trace_events.c
+++ b/kernel/trace/trace_events.c
@@ -15,8 +15,10 @@
 #include <linux/kthread.h>
 #include <linux/tracefs.h>
 #include <linux/uaccess.h>
+#include <linux/bsearch.h>
 #include <linux/module.h>
 #include <linux/ctype.h>
+#include <linux/sort.h>
 #include <linux/slab.h>
 #include <linux/delay.h>
 
@@ -445,6 +447,43 @@ static void ftrace_clear_events(struct trace_array *tr)
 	mutex_unlock(&event_mutex);
 }
 
+static int cmp_pid(const void *key, const void *elt)
+{
+	const pid_t *search_pid = key;
+	const pid_t *pid = elt;
+
+	if (*search_pid == *pid)
+		return 0;
+	if (*search_pid < *pid)
+		return -1;
+	return 1;
+}
+
+static void __ftrace_clear_event_pids(struct trace_array *tr)
+{
+	struct trace_pid_list *pid_list;
+
+	pid_list = rcu_dereference_protected(tr->filtered_pids,
+					     lockdep_is_held(&event_mutex));
+	if (!pid_list)
+		return;
+
+	rcu_assign_pointer(tr->filtered_pids, NULL);
+
+	/* Wait till all users are no longer using pid filtering */
+	synchronize_sched();
+
+	free_pages((unsigned long)pid_list->pids, pid_list->order);
+	kfree(pid_list);
+}
+
+static void ftrace_clear_event_pids(struct trace_array *tr)
+{
+	mutex_lock(&event_mutex);
+	__ftrace_clear_event_pids(tr);
+	mutex_unlock(&event_mutex);
+}
+
 static void __put_system(struct event_subsystem *system)
 {
 	struct event_filter *filter = system->filter;
@@ -777,6 +816,56 @@ static void t_stop(struct seq_file *m, void *p)
 	mutex_unlock(&event_mutex);
 }
 
+static void *p_start(struct seq_file *m, loff_t *pos)
+{
+	struct trace_pid_list *pid_list;
+	struct trace_array *tr = m->private;
+
+	/*
+	 * Grab the mutex, to keep calls to p_next() having the same
+	 * tr->filtered_pids as p_start() has.
+	 * If we just passed the tr->filtered_pids around, then RCU would
+	 * have been enough, but doing that makes things more complex.
+	 */
+	mutex_lock(&event_mutex);
+	rcu_read_lock_sched();
+
+	pid_list = rcu_dereference_sched(tr->filtered_pids);
+
+	if (!pid_list || *pos >= pid_list->nr_pids)
+		return NULL;
+
+	return (void *)&pid_list->pids[*pos];
+}
+
+static void p_stop(struct seq_file *m, void *p)
+{
+	rcu_read_unlock_sched();
+	mutex_unlock(&event_mutex);
+}
+
+static void *
+p_next(struct seq_file *m, void *v, loff_t *pos)
+{
+	struct trace_array *tr = m->private;
+	struct trace_pid_list *pid_list = rcu_dereference_sched(tr->filtered_pids);
+
+	(*pos)++;
+
+	if (*pos >= pid_list->nr_pids)
+		return NULL;
+
+	return (void *)&pid_list->pids[*pos];
+}
+
+static int p_show(struct seq_file *m, void *v)
+{
+	pid_t *pid = v;
+
+	seq_printf(m, "%d\n", *pid);
+	return 0;
+}
+
 static ssize_t
 event_enable_read(struct file *filp, char __user *ubuf, size_t cnt,
 		  loff_t *ppos)
@@ -1334,8 +1423,165 @@ show_header(struct file *filp, char __user *ubuf, size_t cnt, loff_t *ppos)
 	return r;
 }
 
+static int max_pids(struct trace_pid_list *pid_list)
+{
+	return (PAGE_SIZE << pid_list->order) / sizeof(pid_t);
+}
+
+static ssize_t
+ftrace_event_pid_write(struct file *file, const char __user *ubuf,
+		       size_t cnt, loff_t *ppos)
+{
+	struct seq_file *m = file->private_data;
+	struct trace_array *tr = m->private;
+	struct trace_pid_list *filtered_pids = NULL;
+	struct trace_pid_list *pid_list = NULL;
+	struct trace_parser parser;
+	unsigned long val;
+	loff_t this_pos;
+	ssize_t read = 0;
+	ssize_t ret = 0;
+	pid_t pid;
+	int i;
+
+	if (!cnt)
+		return 0;
+
+	ret = tracing_update_buffers();
+	if (ret < 0)
+		return ret;
+
+	if (trace_parser_get_init(&parser, EVENT_BUF_SIZE + 1))
+		return -ENOMEM;
+
+	mutex_lock(&event_mutex);
+	/*
+	 * Load as many pids into the array before doing a
+	 * swap from the tr->filtered_pids to the new list.
+	 */
+	while (cnt > 0) {
+
+		this_pos = 0;
+
+		ret = trace_get_user(&parser, ubuf, cnt, &this_pos);
+		if (ret < 0 || !trace_parser_loaded(&parser))
+			break;
+
+		read += ret;
+		ubuf += ret;
+		cnt -= ret;
+
+		parser.buffer[parser.idx] = 0;
+
+		ret = -EINVAL;
+		if (kstrtoul(parser.buffer, 0, &val))
+			break;
+		if (val > INT_MAX)
+			break;
+
+		pid = (pid_t)val;
+
+		ret = -ENOMEM;
+		if (!pid_list) {
+			pid_list = kmalloc(sizeof(*pid_list), GFP_KERNEL);
+			if (!pid_list)
+				break;
+
+			filtered_pids = rcu_dereference_protected(tr->filtered_pids,
+							lockdep_is_held(&event_mutex));
+			if (filtered_pids)
+				pid_list->order = filtered_pids->order;
+			else
+				pid_list->order = 0;
+
+			pid_list->pids = (void *)__get_free_pages(GFP_KERNEL,
+								  pid_list->order);
+			if (!pid_list->pids)
+				break;
+
+			if (filtered_pids) {
+				pid_list->nr_pids = filtered_pids->nr_pids;
+				memcpy(pid_list->pids, filtered_pids->pids,
+				       pid_list->nr_pids * sizeof(pid_t));
+			} else
+				pid_list->nr_pids = 0;
+		}
+
+		if (pid_list->nr_pids >= max_pids(pid_list)) {
+			pid_t *pid_page;
+
+			pid_page = (void *)__get_free_pages(GFP_KERNEL,
+							    pid_list->order + 1);
+			if (!pid_page)
+				break;
+			memcpy(pid_page, pid_list->pids,
+			       pid_list->nr_pids * sizeof(pid_t));
+			free_pages((unsigned long)pid_list->pids, pid_list->order);
+
+			pid_list->order++;
+			pid_list->pids = pid_page;
+		}
+
+		pid_list->pids[pid_list->nr_pids++] = pid;
+		trace_parser_clear(&parser);
+		ret = 0;
+	}
+	trace_parser_put(&parser);
+
+	if (ret < 0) {
+		if (pid_list)
+			free_pages((unsigned long)pid_list->pids, pid_list->order);
+		kfree(pid_list);
+		mutex_unlock(&event_mutex);
+		return ret;
+	}
+
+	if (!pid_list) {
+		mutex_unlock(&event_mutex);
+		return ret;
+	}
+
+	sort(pid_list->pids, pid_list->nr_pids, sizeof(pid_t), cmp_pid, NULL);
+
+	/* Remove duplicates */
+	for (i = 1; i < pid_list->nr_pids; i++) {
+		int start = i;
+
+		while (i < pid_list->nr_pids &&
+		       pid_list->pids[i - 1] == pid_list->pids[i])
+			i++;
+
+		if (start != i) {
+			if (i < pid_list->nr_pids) {
+				memmove(&pid_list->pids[start], &pid_list->pids[i],
+					(pid_list->nr_pids - i) * sizeof(pid_t));
+				pid_list->nr_pids -= i - start;
+				i = start;
+			} else
+				pid_list->nr_pids = start;
+		}
+	}
+
+	rcu_assign_pointer(tr->filtered_pids, pid_list);
+
+	mutex_unlock(&event_mutex);
+
+	if (filtered_pids) {
+		synchronize_sched();
+
+		free_pages((unsigned long)filtered_pids->pids, filtered_pids->order);
+		kfree(filtered_pids);
+	}
+
+	ret = read;
+	*ppos += read;
+
+	return ret;
+}
+
 static int ftrace_event_avail_open(struct inode *inode, struct file *file);
 static int ftrace_event_set_open(struct inode *inode, struct file *file);
+static int ftrace_event_set_pid_open(struct inode *inode, struct file *file);
 static int ftrace_event_release(struct inode *inode, struct file *file);
 
 static const struct seq_operations show_event_seq_ops = {
@@ -1352,6 +1598,13 @@ static const struct seq_operations show_set_event_seq_ops = {
 	.stop = t_stop,
 };
 
+static const struct seq_operations show_set_pid_seq_ops = {
+	.start = p_start,
+	.next = p_next,
+	.show = p_show,
+	.stop = p_stop,
+};
+
 static const struct file_operations ftrace_avail_fops = {
 	.open = ftrace_event_avail_open,
 	.read = seq_read,
@@ -1367,6 +1620,14 @@ static const struct file_operations ftrace_set_event_fops = {
 	.release = ftrace_event_release,
 };
 
+static const struct file_operations ftrace_set_event_pid_fops = {
+	.open = ftrace_event_set_pid_open,
+	.read = seq_read,
+	.write = ftrace_event_pid_write,
+	.llseek = seq_lseek,
+	.release = ftrace_event_release,
+};
+
 static const struct file_operations ftrace_enable_fops = {
 	.open = tracing_open_generic,
 	.read = event_enable_read,
@@ -1477,6 +1738,26 @@ ftrace_event_set_open(struct inode *inode, struct file *file)
 	return ret;
 }
 
+static int
+ftrace_event_set_pid_open(struct inode *inode, struct file *file)
+{
+	const struct seq_operations *seq_ops = &show_set_pid_seq_ops;
+	struct trace_array *tr = inode->i_private;
+	int ret;
+
+	if (trace_array_get(tr) < 0)
+		return -ENODEV;
+
+	if ((file->f_mode & FMODE_WRITE) &&
+	    (file->f_flags & O_TRUNC))
+		ftrace_clear_event_pids(tr);
+
+	ret = ftrace_event_open(inode, file, seq_ops);
+	if (ret < 0)
+		trace_array_put(tr);
+	return ret;
+}
+
 static struct event_subsystem *
 create_new_subsystem(const char *name)
 {
@@ -2471,6 +2752,9 @@ create_event_toplevel_files(struct dentry *parent, struct trace_array *tr)
 		return -ENOMEM;
 	}
 
+	entry = tracefs_create_file("set_event_pid", 0644, parent,
+				    tr, &ftrace_set_event_pid_fops);
+
 	/* ring buffer internal formats */
 	trace_create_file("header_page", 0444, d_events,
 			  ring_buffer_print_page_header,
@@ -2551,6 +2835,9 @@ int event_trace_del_tracer(struct trace_array *tr)
 	/* Disable any event triggers and associated soft-disabled events */
 	clear_event_triggers(tr);
 
+	/* Clear the pid list */
+	__ftrace_clear_event_pids(tr);
+
 	/* Disable any running events */
 	__ftrace_set_clr_event_nolock(tr, NULL, NULL, NULL, 0);
 
-- 
2.6.1


--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web